Diet of the grey gurnard, Eutrigla gurnardus in the Adriatic Sea, north-eastern Mediterranean

by

Stefano MONTANINI* (1, 2), Marco STAGIONI (1, 2) & Maria VALLISNERI (1)

ABSTRACT. - A total of 449 individuals of grey gurnard Eutrigla gurnardus (Linnaeus, 1758) were collected between December 2007 and July 2008 by bottom trawl surveys in the Adriatic Sea, north-eastern Mediterranean. Stomach contents were investigated in order to analyse diet composition according to fish size and depth. The most important prey were Crustacea, mostly Mysidacea such as Lophogaster typicus and such as rhomboides and Philocheras bispinosus, followed by Teleostei, mostly Perciformes as Gobius niger and Callionymus spp. Dietary differences were found in relation to a critical predator size of 120 mm (TL). Decapoda were more frequently found in the diet of juveniles and Teleostei in that of adults living at a greater depth. Feeding strategy analyses revealed trophic specialization towards reference prey, thus narrowing the trophic niche.

RÉSUMÉ. - Régime alimentaire d’Eutrigla gurnardus (Triglidae) en mer Adriatique (nord-est Méditerranée). Au total, 449 spécimens de grondin gris, Eutrigla gurnardus (Linnaeus, 1758), ont été capturés entre décembre 2007 et juillet 2008 dans le cadre de campagnes de pêche à la traîne en mer Adriatique. Les contenus stomacaux ont été étudiés afin d’analyser leur régime alimentaire, en fonction de la taille et de la profondeur. Les proies les plus importantes sont des Crustacea, surtout des Mysidacea tels que Lophogaster typicus et des Decapoda tels que Goneplax rhomboides et Philo- cheras bispinosus, suivis par des Teleostei, surtout les Perciformes tels que Gobius niger et Callionymus spp. On a constaté une modification du régime alimentaire à partir de la taille critique de 120 mm (LT) : les Decapoda sont plus fréquents dans le régime alimentaire des jeunes et les Teleostei dans celui des adultes vivant à de plus grandes profondeurs. Les résultats de l’analyse du régime alimentaire ont mis en évidence une spécialisation trophique pour des proies de référence et donc une niche trophique étroite.

Key words. - Triglidae - Eutrigla gurnardus - Adriatic Sea - Diet - Feeding habits.

Eutrigla gurnardus (L., 1758) is a widely distributed implementing a multispecies approach to fishery manage- demersal that occurs in the Mediterranean and east- ment (Gulland, 1997). ern Atlantic. This fish inhabits depths of 20-300 m in muddy and sandy substrata (Relini et al., 1999). Its trophic biology in the is poorly known. Current literature MATERIAL AND METHODS only refers to the (Ursin, 1975; Agger and Ursin, 1976, De Gee and Kikkert, 1993, Floeter et al., 2005; Floeter Sampling and Temming, 2005; Weinert et al., 2010) and the Spanish A total of 449 samples of E. gurnardus were collected (Catalan) Mediterranean coast (Moreno-Amich, 1994). As seasonally by international bottom trawl surveys in the Med- to Italian coastal areas, available data are very limited, dat- iterranean Sea in winter and summer at depths between 31 ing back several decades (Valiani, 1934). and 185 m in the northern-middle Adriatic Sea from the Gulf The present study sets out to analyse stomach contents of Trieste (45°40’N; 13°37’E) to the line joining the Gargano and dietary changes in relation to abiotic (depth) and biotic to Dubrovnik (42°08’N; 15°16’E). The bottom trawl surveys (length) variables with a view to better understanding the were carried out from December 2007 to March 2008 and ecological role of this species in Adriatic demersal marine from June to July 2008 along the Italian coasts (Fig. 1). communities similar to those in the North Sea where the spe- Total length (TL, mm) and weight (W, g) of each individ- cies is considered to be an emerging key predator that feeds ual were measured in the laboratory to an accuracy of 0.1 g. on invertebrates as well as on fish (Floeteret al., 2005; Wein- Sex was determined on the basis of macroscopic gonads, and ert et al., 2010). Knowledge of the feeding ecology of both specimens were classified as females (F), males (M), juve- commercial and non-commercial fish species is essential for niles (J, macroscopically unidentifiable sex) or not deter-

(1) Department of Evolutionary Experimental Biology, University of Bologna, Via Selmi 3, 40126 Bologna, Italy. [[email protected]] [[email protected]] (2) Laboratory of and Fishery, University of Bologna, Viale Adriatico 1/N, 61032 Fano (PU), Italy. * Corresponding author [[email protected]]

Cybium 2010, 34(4): 367-372. Feeding habits of Eutrigla gurnardus in the Adriatic Sea Mo n t a n i n i e t a l .

prey (Chrisafiet al., 2007). These indexes were used to cal- culate a fullness index (FI = number of stomachs with the same degree of fullness / total number of stomachs examined x 100) and to determine changes in trophic level according to two main size classes correlated to depth. Group average hierarchical cluster analysis was carried out using the Bray-Curtis similarity index (Clarke and War- wick, 1994) to assess differences between predator length and number (%N) of ingested food categories, data being transformed by Wisconsin double standardization to miti- gate prey item range. This analysis was made to determine ontogenetic changes in fish diet. Specimens were subdivided into 6 size-classes. General feeding strategy was analysed according to the Costello method (1990) as revised by Amundsen et al. (1996). The diagram generated by this method provides a two-dimensional representation in which each prey point is obtained by plotting the frequency of occurrence (%O) Figure 1. - Sampled sites in the north Adriatic Sea, north-eastern Mediterranean. against prey-specific abundance (PSA) calculated according to the following formula: Pi = (∑Si / ∑Sti) x 100 where Si is the stomach content (weight) of prey item i and Sti is the mined (ND, individual not examined) (Relini et al., 2008). total stomach content (weight) of only those predators with 156 females, 118 males, 109 juveniles and 66 undetermined prey item i in their stomach. Prey importance and predator were recorded. The stomachs of each of these specimens feeding strategy were determined on the basis of dot distri- were removed and preserved in 70% ethanol for subsequent bution along diagram diagonals and axes. examination in the laboratory.

Diet analyses RESULTS Food items were identified to the lowest taxonomic level possible, taking into account each prey type and digestion General diet composition state. Food items were counted, weighed (nearest 0.001 g) Five taxa (Tab. I) and namely Arthropoda (represented by and photographed by means of a NIKON P5100 digital Crustacea), Chordata (represented by Teleostei), Mollusca, camera connected to a Leica MZ6 stereomicroscope. Inor- Nematoda and Echinodermata, and 33 prey items (Tab. II), ganic particles and detritus present in the stomachs were also were identified in this study. The diet of E. gurnardus was recorded. Data were collected in a database and processed mostly made up of (%N = 94.2; %W = 71.9), and with R software, version 2.10.1. to a lower extent of teleost fishes (%N = 2.6; %W = 27.5) The relative importance of each item was calculated and molluscs (%N = 2.3; %W = 0.3). Other prey were nema- on the basis of item quantity (number) divided by the total todes (%N = 0.5; %W < 0.1) and echinoderms (%N = 0.1; number of overall prey (%N = number of prey i / total %W < 0.1). Teleost fishes were more important in terms of number of prey x 100), weight proportion (%W = weight of prey i / total weight of overall prey x 100) and frequency of occurrence (%O = number of stomachs containing prey Table I. - Prey taxa recorded in the stomach content of Eutrigla gur- i / total number of stomachs containing prey i x 100). These nardus from the Adriatic Sea, all individuals merged. %N = numer- ical composition; %W = weight proportion; %O = frequency of values were used to calculate the relative importance index occurrence; IRI = index of relative importance; %IRI = IRI propor- (IRI): IRI = %O (%N + %W) and IRI proportion (%IRI). tion. Stomach content coefficients were determined according Prey Taxa %N %W %O IRI %IRI to an empirical scale as empty (A) = stomach barely bloat- ed with no and/or with only a few small prey; semi-empty Arthropoda 94.15 71.93 94.04 15618.6 97.61 (B) = stomach slightly bloated with only a few and/or with Chordata 2.57 27.49 11.91 358.05 2.24 considerably small prey; semi-full (C) = stomach sufficient- Mollusca 2.25 0.34 8.46 21.94 0.14 ly bloated with considerable amounts of small prey or with nd 0.51 0.24 2.51 1.9 0.01 only a few large prey; full (D) = stomach massively bloated Nematoda 0.45 0 1.57 0.71 0 with large amounts of small prey or with only a few large Echinodermata 0.06 0 0.31 0.02 0

368 Cybium 2010, 34(4) Mo n t a n i n i e t a l . Feeding habits of Eutrigla gurnardus in the Adriatic Sea

Table II. - Diet composition of Eutrigla gurnardus from the Adriatic Sea, all individuals weight proportion than of occurrence fre- merged. %N = numerical composition; %W = weight proportion; %O = frequency of quency and proportion (Fig. 2). occurrence; IRI = index of relative importance; %IRI = IRI proportion. Decapod crustaceans were the most Prey species %N %W %O IRI %IRI important prey items of this population Mysidacea of E. gurnardus. Amongst Decapoda, the dominant species in order of importance Lophogaster typicus 31.73 24.72 21.94 1238.58 50.49 were: Goneplax rhomboides, Philocheras Acanthomysis longicornis 6.81 0.48 3.76 27.42 1.12 bispinosus, Liocarcinus spp., Solenocera Gastrosaccus sanctus 0.5 0.03 0.63 0.33 0.01 membranacea, Alpheus glaber. Lophog- Gastrosaccus sp. 0.17 0.03 0.31 0.06 0 aster typicus (Mysidacea) appeared to be Decapoda: Brachyura the most abundant food in the stomachs Goneplax rhomboides 21.59 22.06 20.69 903.19 36.82 examined. Amongst teleost fishes, the Liocarcinus sp. 7.14 2.93 7.21 72.59 2.96 most preyed species were Gobius niger Liocarcinus depurator 2.16 6.81 3.45 30.92 1.26 and Callionymus spp. Decapoda: Macrura-Natantia Fullness index Philocheras bispinosus 10.13 1.7 6.27 74.2 3.02 E. gurnardus fullness index varied Alpheus glaber 1.99 6.06 3.45 27.77 1.13 significantly for the two size classes Philocheras sp. 5.15 0.78 3.76 22.29 0.91 calculated as the smaller and the larger Solenocera membranacea 3.99 2.51 3.45 22.39 0.91 than the median, the latter being 120 mm 2 Processa spp. 2.66 0.21 1.57 4.49 0.18 (chi = 12.7752, df = 3, p < 0.05) (Fig. 3). Pontophilus sp. 0.17 1.34 0.31 0.47 0.02 Stomach fullness was greater in adults Pontophilus spinosus 0.17 0.99 0.31 0.36 0.02 living at depths greater than the juve- niles (chi2 = 26.0871, df = 6, p < 0.001). Philocheras echinulatus 0.17 0.12 0.31 0.09 0 Adults were found to feed on larger Decapoda: Anomura prey (teleost fish and decapods such as Galathea sp. 0.66 0.18 0.63 0.53 0.02 ) with a greater leathery content, Galathea intermedia 0.66 0.12 0.63 0.49 0.02 related to increased energy requirements Munida spp. 0.33 3.28 0.31 1.13 0.05 (Fig. 4). Sex-wise, greater fullness was Decapoda: Macrura-Reptantia recorded for females than for males 2 Callianassa spp. 0.17 0.24 0.31 0.13 0.01 (chi = 12.9224, df = 3, p < 0.05). Jaxea nocturna 0.17 0.21 0.31 0.12 0 Diet variation with fish size Teleostei The diet similarity dendrogram (Fig. Gobius niger 0.83 10.54 1.57 17.82 0.73 5) reveals a first dichotomy marking out Callionymus sp. 0.5 2.72 0.94 3.02 0.12 two distinct main groups depending on Arnoglossus laterna 0.17 9.31 0.31 2.97 0.12 total length (TL) being greater or less Callionymus maculatus 0.17 2.36 0.31 0.79 0.03 120 mm. Other dichotomies were found, Bivalvia marking out a further six groups accord- Tellina spp. 0.33 0.09 0.63 0.26 0.01 ing to total fish length, and namely class I containing fish of up to 80 mm; class II Mactra corallina 0.17 0.03 0.31 0.06 0 from 81 to 100 mm; class III from 101 to Flexopecten glaber proteus 0.17 0 0.31 0.05 0 120 mm; class IV from 121 to 140 mm; Hyalopecten similis 0.17 0 0.31 0.05 0 class V from 141 to 160 mm, and class Opisthobranchia VI longer than 160 mm. Juveniles (class Turritella communis 0.33 0.06 0.63 0.25 0.01 I; class II) were active predators since Odostomia acuta 0.17 0 0.31 0.05 0 their first life stages. Their predation Euphausiacea was directed at smaller crustaceans such Meganyctiphanes norvegica 0.17 0.12 0.31 0.09 0 as Philocheras bispinosus and Proc- essa spp. Specimens in class III were Amphipoda found to start to feed on some molluscs Ampelisca spp. 0.17 0 0.31 0.05 0 but the most significant change in feed- Leucothoe spp. 0.17 0 0.31 0.05 0 ing behaviour was observed for speci-

Cybium 2010, 34(4) 369 Feeding habits of Eutrigla gurnardus in the Adriatic Sea Mo n t a n i n i e t a l .

Figure 4. - Variations in stomach fullness of E. gurnardus in rela- tion to sex. Juveniles and adults (females and males).

Figure 2. - Proportion of main prey taxa in the stomachs of E. gur- nardus, all individuals merged. %N = numerical composition; %W = weight proportion; %O = frequency of occurrence; nd = not determined.

Figure 5. - Dendogram showing feeding similarities among six E. gurnardus size classes from the Adriatic Sea. Class I (60- 80 mm); Class II (80-100 mm); Class III (100-120 mm); Class VI (120-140 mm); Class II (140-160 mm); Class III (160-260 mm). Species diversity Bray-Curtis similarity index (cluster method: average).

such as amphipods, bigger decapods such as Brachyura and Alpheus glaber, mysids, and molluscs. Specimens in these Figure 3. - Variations in stomach fullness of E. gurnardus in relation classes were also found to begin to prey necto-benthic items to two size classes. Class I (60-120 mm); Class II (120-255 mm). such as Callyonimus maculatus. Specimens of more than 160 mm (class VI), although preying crustaceans such as mens in classes IV and V. The variety of food items eaten Pontophilus spp. and Liocarcinus depurator, were found to by grey gurnards was seen to increase during life stages cor- mainly exhibit piscivorous feeding habits. Their prey were responding to classes IV and V, when the highest value of seen to include Callyonimidae, Gobius niger and Pleuronec- niche breadth was attained. Preys include benthic food items tiformes such as Arnoglossus laterna.

370 Cybium 2010, 34(4) Mo n t a n i n i e t a l . Feeding habits of Eutrigla gurnardus in the Adriatic Sea

Feeding strategy Prey-wise, feeding strategy analysis on the basis of stom- ach contents revealed crustaceans Lophogaster typicus and Goneplax rhomboides to be the most dominant prey for the predator populations considered (see top right corner of the graph). Such feeding strategy is indicative of an over- all predator population with a narrow feeding niche (Fig. 6). Despite their low occurrence, the teleost Gobius niger and the crustaceans Liocarcinus depurator, Alpheus glaber and Solenocera membranacea were observed with great specific abundance. These more palatable preys were consumed by only a few individuals displaying high specialization (see upper left corner of the graph). Smaller crustaceans such as Acanthomysis longicornis, Philocheras spp., Processa spp. and teleost fishesCallionymus spp. represented rare and less important prey (see lower left corner of the graph). Graphical feeding strategy analysis shows inter-phe- notype contribution to be quite high (high BPC), with a fair number of the majority of prey points being positioned Figure 6. - Prey-specific abundance plotted against frequency of toward the upper left corner of the graph. This points to the most occurring prey items in the diet of E. gurnardus from the Adriatic Sea. %O = frequency of occurrence; PSA (%) = prey-spe- fact that a number of individuals of the predator population cific abundance. of E. gurnardus considered were specialized in the predation of different resource types, reflecting dietary changes over the lifecycle of this species. Agger and Ursin, 1976; De Gee and Kikkert, 1993; Moreno- Amich, 1994; Floeter et al., 2005; Floeter and Temming, 2005). Ontogenetic body increase allows growing predators DISCUSSION to successfully ingest more elusive, larger prey (Karpouzi and Stergiou, 2003). In addition, adults move faster and The present study bears out the importance of investigat- are therefore capable of feeding on prey with higher motil- ing dietary habits, predator-prey relationships and trophic ity (i.e., fish). They become more opportunistic, apparently levels in order to quantify the ecological role of the vari- engaging in a more generalist-oriented foraging behaviour. ous components of any marine community. Ideally, priority According to the plotted pattern, smaller crustaceans were should be given to studying the trophic biology of scantly seen to be important prey for the smaller sized classes of the known stocks such as, for instance, E. gurnardus, especially predator, while larger crustaceans and fishes were preferred when it is considered that they may limit or affect the trophic by larger individuals. niches of other species of commercial importance. The critical size at which E. gurnardus in the area Very few studies have been carried out on the feeding examined was found to switch over from feeding mainly habits of the Mediterranean grey gurnard particularly in Ital- on crustaceans to fishes, to predate at greater depths and to ian seas (Valiani, 1934). On the basis of stomach content start reaching sexual maturity, was observed to be at around analysis, some authors have reached the conclusion that this 120 mm TL. The differentiation between females and males predator feeds on crustaceans (mainly Decapoda natantia and starts at lengths close to and less than 100 mm TL. 50% of Mysidacea) and teleost fishes (Reys, 1960 - Gulf of Lions; males have been found to attain maturity at around 120 mm Moreno-Amich, 1994 - Catalan coast). Molluscs were also TL, while 50% of females at around 150 mm TL (Vallisneri observed to be present in the stomachs, but were deemed to et al., unpublished data). As to size at maturity, our findings contribute only moderately to the overall diet. are comparable to those reported in the literature, regardless Our findings based on the study of Adriatic Sea popula- of location (Papaconstantinou, 1982 - Greece; Dorel, 1986 tions of the predator are in line with those of these authors. - Gulf of Biscay, France; Muus and Nielsen, 1999 - Den- Grey gurnards are epibenthic feeders preying mainly crus- mark). Minor differences as were observed may be attribut- taceans. Depending on lifecycle stage, it was observed that ed to local conditions, especially temperature that stimulates E. gurnardus increases its predation of fish. Such behav- sexual maturation, and to the study methodology adopted by iour probably depends on the predator’s changing ener- the different authors. getic requirements with growth as well as on its changing Stomach fullness is greater in females than in males, a predation capacities in relation to fish size (Ursin, 1975; fact that may be accounted for by the higher fitness costs

Cybium 2010, 34(4) 371 Feeding habits of Eutrigla gurnardus in the Adriatic Sea Mo n t a n i n i e t a l .

(i.e., big eggs and large fecundity) incurred by females and DOREL D., 1986. - Poissons de l’Atlantique nord-est : Relations by their access to the best spawning sites. Taille-Poids. 165 p. Nantes, France : Institut francais de Recher- che pour l’Exploitation de la Mer. The analysis of prey-specific abundance against frequen- FLOETER J. & TEMMING A., 2005. - Analysis of prey size pref- cy of occurrence shows that E. gurnardus has a narrow niche erence of North Sea whiting, saithe, and grey gurnard. ICES J. width, Lophogaster typicus and Goneplax rhomboides being Mar. Sci., 62: 897-907. the dominant prey. The most abundant food in their stomachs FLOETER J., KEMPF A., VINTHER M., SCHRUM C. & TEM- was found to be represented by the species Lophogaster typi- MING A., 2005. - Grey gurnard (Eutrigla gurnadus) in the cus (Mysidacea); this finding is in line with that of the litera- North Sea: an emerging key predator? Can. J. Fish. Aquat. Sci., 62: 1853-1864. ture relating to the stomach contents of other species of the GULLAND L.A., 1977. - Goals and objectives of fishery manage- same Triglidae family in the south Mediterranean (Terrats et ment. FAO Fish. Tech. Rep., 166. 14 p. Rome: FAO. al., 2000). It is also in line with the results reported by Collo- KARPOUZI V.S. & STERGIOU K.I., 2003. - The relationships ca et al. (2010) for other Mediterranean gurnards (Aspitrigla between mouth size and shape and body length for 18 species cuculus, Lepidotrigla cavillone and L. dieuzeidei) that were of marine fishes and their trophic implications. J. Fish Biol., found to be characterized by a specialist trophic behaviour. 62: 1353-1365. A significant negative correlation between prey type similar- MONTANINI S., STAGIONI M. & VALLISNERI M., 2008. - Note sulla biologia di Chelidonichthys lucernus (Teleostei: Trigli- ity and depth distribution similarity was also observed for dae) nell’Alto-Medio Adriatico. Biol. Mar. Medit., 15: 340- Mediterranean demersal fish, with gurnards overlapping in 341. both food and depth preferences (Colloca et al., 2010). MORENO-AMICH R., 1994. - Feeding habits of grey gurnard, In line with the literature (Floeter et al., 2005; Montanini Eutrigla gurnardus (L., 1758), along the Catalan coast (north- et al., 2008), the Adriatic population studied by us shows that western Mediterranean). Hydrobiologia, 273: 57-66. E. gurnardus potentially competes as a predator with other MUUS B.J. & NIELSEN J.G., 1999. - Sea Fish. Scandinavian Fish- ing Year Book. 340 p. Denmark: Hedehusene. fish species. As reported by Floeter and Temming (2005), in PAPACONSTANTINOU C., 1982. - Age and growth of grey gur- the North Sea grey gurnard predation is in fact a very critical nard (Eutrigla gurnardus L. 1758) in the Pagassitikos Gulf process with a significant top-down effect on whiting (Mer- (Greece). Inv. Pesq., 46: 191-213. langius merlangus) and potentially also on recruitment. REYS J.P., 1960. - Étude de la nourriture de quelques poissons démersaux du golfe du Lion. Rec. Trav. Stat. Mar. Endoume, 20: 65-97. Acknowledgements. - We thank Pr Corrado Piccinetti for his assistance in bottom trawl surveys and his valuable suggestions. Relini G., Bertrand J. & Zamboni A., 1999. - Synthesis of the knowledge on bottom fishery resources in Central Mediter- ranean (Italy and Corsica). Biol. Mar. Medit., 6: 470-474. REFERENCES RELINI G., CARPENTIERI P. & MURENU M., 2008. - Medits Instruction Manual. Biol. Mar. Medit., 15: 31-34. AGGER P. & URSIN E., 1976. - Food and food preference of the TERRATS A., PETRAKIS G. & PAPACONSTANTINOU C., grey gurnard. ICES Doc., C.M./F, 41: 1-8. 2000. - Feeding habits of Aspitrigla cuculus (L., 1758) (red AMUNDSEN P.A., GABLER H.M. & STALDVIK F.J., 1996. - A gurnard), Lepidotrigla cavillone (Lac., 1802) (large scale gur- new approach to graphical analysis of feeding strategy from nard) and Trigloporus lastoviza (Brunn., 1768) (rock gurnard) stomach contents data modification of the Costello (1990) around Cyclades and Dodecanese Islands (E. Mediterranean). Medit. Mar. Sci. method. J. Fish Biol., 48: 607-614. , 1/1: 91-104. URSIN E., 1975 - Food and food preference of the grey gurnard. CHRISAFI E., KASPIRIS P. & KATSELIS G., 2007. - Feeding ICES Doc. CM/F habits of sand smelt (Atherina boyeri, Risso 1810) in Trichonis , 4a: 8 p. Lake (Western Greece). J. Appl. Ichthyol., 23: 209-214. VALIANI S., 1934. - Contributo allo studio dell’alimentazione dei T. lyra, T. gurnardus Boll. Pesca Piscic. Idrobiol CLARKE K.R & WARWICK R.M., 1994. - Change in Marine pesci, . ., 10: Communities: an Approach to Statistical Analysis and Interpre- 37-46. tation. 144 p. Natural Environment Research Council (UK). WEINERT M., FLOETER J., KRONCKE I. & SELL F., 2010. - COLLOCA F., CARPENTIERI P., BALESTRI E. & ARDIZZONE The role of prey composition for the condition of grey gurnard (Eutrigla gurnardus). J. Appl. Ichthyol., 26: 75-84. G., 2010. - Food resource partitioning in a Mediterranean demersal fish assemblage: the effect of body size and niche width. Mar. Biol., 157: 565-574. COSTELLO M.J., 1990. - Predator feeding strategy and prey importance: a new graphical analysis. J. Fish Biol., 36: 261- 263. DE GEE A. & KIKKERT A.H., 1993. - Analysis of the grey gur- nard (Eutrigla gurnardus) samples collected during the 1991. International Stomach Sampling Project. ICES Doc., C.M./G, Reçu le 21 juin 2010. 14: 1-25. Accepté pour publication le 16 novembre 2010.

372 Cybium 2010, 34(4)